I've voted "other" - and I'd like to take the opportunity to explain why.
People think about the outline, shading, colour/values, concept, and details - but one area that is frequently neglected is
the big relationships of the shapes, yet I think this is the one aspect which more generally determines the difference between the professional/accomplished drawing and that of the beginner than any other. I'm not just talking about formulaic "proportions" here, but about getting the big shapes and forms right, for the specific pose/scene.
Here's an article I wrote on this issue for another site:
Spatial frequency, composition, and likenessI'd like to introduce, and discuss, the concept of spatial frequency, in relation to art.

The dot on the left here shows a step in value at its edge which abruptly goes from white to black - the is described as "high spatial frequency"
The dot on the right, however, changes in a way which is much more gradual and spread over a much greater distance, and area - such a feature is described as "low spatial frequency".
Now we can use this concept to think about several issues of concern to artists, such as:
- Painterly vs Tight
- Line vs Mass
- Composition
- Proportions and likeness.
I see a lot of people concentrating on detail, and then wondering why their painting has poor composition, or why they have failed to capture likeness. I suspect that this is often because they are not aware of the importance of working in, and thinking about, the low spatial frequency range.
If we look at Ingres "Madame Moitessier" (1856), it is typical of a "classical" painting style, being "tight", with lots of fine linear detail

- but he also gets the
low-spatial frequency information correct too - the painting is still striking - has a good composition - when seen from a distance, and blurred:

Nature is fractal - it shows a similar level of intricacy at different scales - we turn the magnification up, and the level of intricacy appears to remain constant.
We thus are attuned naturally to find attractive, images which replicate this scaling phenomenon, at least across some range of scales.
This is why some paintings look interesting from across the room, even before we can make out what they actually are. This can be thought of in terms of being good abstract composition.
Here's another painting with that "across the room" thing going for it:

Sargent's "Wertheimer Children" 1902 - again, it's got a bold composition...
and when we get closer, we see

but if we get closer than the ideal viewing distance from these two paintings, we see something interesting in the comparison:

because we can see that the Ingres is painted in precise minute painstaking detail, while the Sargent turns out to be constructed from a series of bold sweeping brushstrokes, with very little in the way of detail.
The "promise" of the next level of detail becoming apparent as we get closer, pulls us in towards a painting - and with a "tight" painting like the Ingres, we can get a lot closer than the "optimal viewing distance for the whole painting" and still keep resolving new detail (like the real world).
However, for the Sargent, we discover that as we get closer than the "optimal", those areas which looked entirely consistent with containing more detail turn out instead to break down into blobs and splotches -
which is surprising - and is what is generally called "painterly" - what Sargent has done is approach the image fundamentally from the low-spatial-frequency end, and pushed the spatial frequencies up
just enough to be convincing at the optimal viewing distance, or slightly closer. Now this approach can be seen as very efficient, because Sargent gives us just enough to be convincing, without spending a lot of time on painting detail which is unnecessary - however, it
works because Sargent has a very solid understanding of this aspect of "the big picture" - low spatial frequency.
This also relates to soft edges, and differential focus in relation to retinal visual acuity. (The Sargent better mimics the way we see, whereas the Ingres is hyperreal).
It's probably a reason why those who
draw in tight detail, or with line, sometimes have difficulties in adapting to painting - as they are approaching it from the outset with a high-spatial-frequency mindset.
One might also add that this approach gives some insight onto why abstract painting is sometimes boring, as it's insufficiently fractal - one gets closer, and finds there's nothing new to see, after only one or two levels.
One of the ideas that I'm trying to get across here is that the ideal image has data across a wide range of the spatial frequency spectrum, but that it's the middle to low end where people perhaps ought to be concentrating their efforts - this range is
more important to composition:

and to general proportions and likeness:

than is high spatial frequency (or in more usual artistic terminology, fine detail).
Now part of this is getting away from the idea that the large-scale stuff is just a "rough" preparation for the fine detail - but that the earliest stages of composition, proportions, and block-in require just as much,
if not more attention (and a preparedness to stand well back) than the final touches of (high-spatial frequency) detail.
Dave